Plagioclase Megacrysts in Mesoproterozoic Amphibolites from the New Jersey Highlands, USA: Indicators of Mixed-Source Magma and Fractionation Interruption in Anorthosite-Dominated Terrains

Abstract

Rare amphibolite in the New Jersey Highlands containing plagioclase megacrysts ≤13 cm long forms bodies 0.5 to 2 m thick that preserve a penetrative metamorphic fabric and have sharp, conformable contacts against Mesoproterozoic country rocks. The megacrystic amphibolites were emplaced as thin dikes along extensional faults between 1160 and 1130 Ma. Amphibolites contain weakly zoned, subhedral andesine megacrysts (An_29–44) in a matrix of plagioclase (An_18–38), magnesio-hastingsite, biotite, diopside, Fe-Ti oxides, and apatite. The whole-rock major oxide composition of the megacrystic amphibolite matrix has high TiO₂ (2.07 wt.% ± 2.0%), Al₂O₃ (17.03 wt.% ± 0.87%), and Fe₂O₃^t (12.84 wt.% ± 3.2%) that represents an Al-Fe-rich mafic magma type that is characteristic of anorthosite associations globally. The whole-rock, rare earth element (REE) composition of the megacrystic amphibolite matrix is characterized by enrichments in light rare earth elements (LREEs) (La/Yb_N = 1.73–10.69) relative to middle (MREEs) and heavy (HREEs) rare earth elements (Gd/Yb_N = 1.30–1.85), and most samples have small positive Eu anomalies (Eu/Eu* = 0.95–1.25). The megacrystic amphibolite matrix is also enriched in large ion lithophile elements (LILEs) and depleted in high field strength elements (HFSEs) (e.g., Ba/Nb = 24–22). Megacrystic amphibolites formed through partial melting of subduction-modified lithospheric mantle that fractionated olivine and plagioclase, producing a high-Al-Fe mafic magma. Plagioclase megacrysts formed through extraction of a plagioclase-rich crystal-liquid mush from anorthosite that mixed with mafic magma and collected in the upper part of the mid-crustal magma (depth of ~20 km based on Al-in-hornblende geobarometry) chamber through flotation. Periodic tapping of this mixed source by extensional fractures led to emplacement of the amphibolites as dikes and may have interrupted the extensive fractionation and plagioclase separation necessary to form voluminous anorthosite intrusions.

1. Introduction

Mafic intrusive rocks that contain plagioclase megacrysts are recognized globally and include occurrences that are associated with anorthosite, such as mafic dikes in Archean anorthosite complexes in Canada [1] and gabbroic dikes in the Gardar Province, south Greenland [2], or that lack an association with anorthosite, such as amphibolites in the Flekkefjord complex, south Norway [3], basalts in the Deccan Volcanic Province, India [4], in the Bandera lava field, New Mexico, USA [5], the alkaline rock province of southwest Japan [6], and the Kurile Island Arc, northwest Pacific Ocean [7]. In addition, mafic dikes and/or mafic intrusions that lack plagioclase megacrysts and are compositionally similar to the megacrystic amphibolites in this study and associated with anorthosite occur in the Eastern Granulites of Tanzania [8], Western Sierras Pampeanas in Argentina [9], Damian gabbro–anorthosite suite in China [10], and Laramie anorthosite complex in Wyoming, USA [11]. Other occurrences of mafic intrusions that are associated with anorthosite but lack plagioclase megacrysts occur in Grenvillian terranes in eastern North America, most notably in the Adirondack Highlands [12,13,14,15], and at numerous locations in the Canadian Grenville Province ([16] and references therein). These mafic intrusions are characterized by high Al-Fe compositions (e.g., [13]) and commonly display sharp, discordant, and intrusive contacts with spatially associated anorthosite [12].

Many occurrences of mafic intrusions in the Adirondack Highlands and the Grenville Province were emplaced as part of anorthosite–mangerite–charnockite–granite (AMCG) suites during magmatic intervals at ca. 1165–1140 Ma [17] in the former and ca. 1160–1140 Ma and 1080–1010 Ma [18,19] in the latter. South of the Adirondack Highlands, the ages of anorthosite and associated rocks generally remain poorly constrained, although separate bodies of anorthosite from Grenvillian inliers in eastern Virginia yielded U-Pb zircon ages of 1045 ± 10 Ma [20] and 1011 ± 3.5 and 1010 ± 2.2 Ma [21] that fall within the younger pulse of AMCG magmatism. Much less is known about AMCG-type magmatism in Grenville-age inliers south of the Adirondacks, except that anorthosite and contemporaneous mafic rocks are not recognized in the Green Mountains or the Berkshire Mountains and are present in only sparse amounts in the New York Hudson Highlands, New Jersey Highlands, and Honey Brook Upland in Pennsylvania (Figure 1). Studies of anorthosite and spatially associated mafic rocks in these Appalachian inliers are few [22,23,24,25], and the geologic relationships between them are poorly understood. It is worth noting that to our knowledge plagioclase megacrysts are not reported in other occurrences of mafic rocks north of the study area, and therefore they may be unique to the New Jersey Highlands.

We present the first results from integrated datasets for megacrystic amphibolites from the New Jersey Highlands in the Grenville of the central Appalachians that include a detailed mineralogical analysis of the major phases present, major and trace element geochemical analysis, oxygen isotope analysis of plagioclase, and an assessment of the geologic relationships based on field observations. The results place reasonable constraints on a suitable source for the amphibolites, magmatic conditions during their formation, and their geodynamic setting of emplacement. In addition, we examine possible processes for the formation of the megacrysts that are consistent with the available data.

2. Geologic Setting

The New Jersey Highlands constitute one of numerous inliers that extend along eastern North America that contain rocks of Grenvillian age (Figure 1). Most of the 1000 km² area of the New Jersey Highlands is underlain by rocks of Mesoproterozoic age that are separated into western and eastern parts (Figure 2) by downfaulted Paleozoic rocks of the Green Pond Mountain region.

Mesoproterozoic rocks of the Highlands include a heterogeneous assemblage of lithotectonic units that record a complex evolution. The calc-alkaline Wanaque tonalite gneiss (U-Pb zircon ages of 1366 ± 9 to 1363 ± 17 Ma) and gneissic rhyolite, dacite, tonalite, diorite, and amphibolite of the Losee Suite (U-Pb zircon ages of 1282 ± 7 to 1248 ± 12 Ma) are interpreted as having formed in a continental magmatic arc above a west-dipping subduction zone outboard of the eastern Laurentian margin [26,27]. These plutonic arc rocks are spatially associated with a thick sequence of supracrustal rocks that include bimodal felsic and mafic metavolcanic rocks, quartzofeldspathic and calc-silicate paragneiss, and marble that formed in a back-arc basin inboard of the magmatic arc [26]. Of these, felsic metavolcanic rocks from different stratigraphic intervals yielded U-Pb ages of 1299 ± 8 to 1238 ± 22 Ma [27], indicating that supracrustal rocks are coeval with 1282–1248 Ma arc-related magmatism of the Losee Suite.

Termination of arc-related magmatism and closure of the back-arc basin took place by ca. 1.2 Ga during west-directed orogenic convergence. Widespread and abundant post-subduction-related magmatism consists of ferroan, metaluminous, A-type hastingsite and hedenbergite granitoids that yielded U-Pb zircon ages of 1188 ± 5 to 1182 ± 11 Ma [27]. Late- to post-orogenic magmatism is dominated by metaluminous A-type magmatism that includes weakly deformed syenogranite (U-Pb zircon age of 1027 ± 6 Ma) and the undeformed Mount Eve Granite (U-Pb zircon age of 1019 ± 5 Ma) [27]. The Mount Eve is characterized by sharp, discordant contacts and xenoliths of foliated country rocks [28,29].

Mesoproterozoic rocks in the New Jersey Highlands were metamorphosed to upper amphibolite- to granulite facies conditions during the Ottawan phase of the Grenvillian Orogeny. The timing of this high-grade metamorphism is constrained by metamorphic overgrowths on zircon and monazite that yielded SHRIMP U-Pb ages of 1045 ± 6 to 1024 ± 7 Ma [27]. The penetrative NE-trending, SE-dipping planar metamorphic fabric and dominant northwest-verging isoclinal folds in the Mesoproterozoic rocks were formed during orogenic convergence and are therefore older than 1024 Ma. Emplacement of the Mount Eve Granite at 1019 Ma places a firm temporal constraint on the close of compressional orogenesis in the New Jersey Highlands and contiguous areas following Ottawan continental collision.

Estimates of temperature and pressure during Ottawan high-grade metamorphism are in reasonably good agreement from both the western and eastern Highlands. Peak metamorphic temperatures reached ~780 °C based on various geothermometers ([26] and references therein) and averaged 769 °C based on regional calcite–graphite geothermometry [30]. Estimates of the confining pressure range from 410 to 710 MPa, with most estimates clustering between 550 and 600 MPa [26]. The data from the New Jersey Highlands are consistent with peak metamorphic conditions of low-pressure, high-temperature granulite facies. These data provide little evidence for a metamorphic discontinuity between the western and eastern parts of the region, and based on the regional commonality of pressure and temperature conditions and lithologies, faults partitioning the region into structural panels lack evidence for an origin as tectonic boundaries separating disparate terranes [26].

3. Analytical Methods

Electron microprobe analysis (EMPA) of the major phases in the megacrystic amphibolites was performed at Rutgers University on a JEOL JXA-8600 Superprobe, and the data are presented in

Table 1. Representative EMPA of plagioclase.

Sample	TQ423-3			TQ423-4	EAS2		BLR205		ST106	DOV189
Note	Mega Core	Mega Rim	Matrix	Matrix	Mega Rim	Matrix	Mega Rim	Matrix	Matrix	Matrix
	n = 5	n = 2	n = 38	n = 9	n = 4	n = 8	n = 6	n = 17	n = 41	n = 10
Oxides in wt.%
SiO2	56.51	56.09	57.93	60.03	59.29	60.43	61.02	59.55	56.53	63.70
TiO2	0.00	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.02
Al2O3	27.43	27.80	27.17	25.58	25.19	25.57	25.22	25.25	27.21	23.18
FeOt	0.08	0.08	0.11	0.13	0.12	0.20	0.03	0.10	0.10	0.06
MnO	0.01	0.00	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.00
MgO	0.00	0.01	0.04	0.00	0.01	0.09	0.01	0.01	0.01	0.01
CaO	8.75	9.21	8.19	6.77	6.53	6.55	6.11	6.38	8.43	3.90
Na2O	6.60	6.30	7.12	7.85	8.21	8.48	7.97	7.88	7.06	9.48
K2O	0.25	0.23	0.25	0.30	0.40	0.38	0.42	0.36	0.17	0.38
Total	99.63	99.73	100.83	100.68	99.77	101.72	100.80	99.55	99.53	100.73
Structural formulae based on 8 oxygens
Si	2.547	2.527	2.576	2.661	2.658	2.658	2.694	2.668	2.551	2.798
Ti	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.001
Al	1.457	1.476	1.424	1.337	1.331	1.326	1.312	1.333	1.447	1.200
Fe	0.003	0.003	0.004	0.005	0.004	0.007	0.001	0.004	0.004	0.002
Mn	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Mg	0.000	0.001	0.003	0.000	0.001	0.006	0.001	0.001	0.001	0.001
Ca	0.423	0.445	0.390	0.322	0.314	0.309	0.289	0.306	0.408	0.184
Na	0.577	0.550	0.614	0.675	0.714	0.723	0.682	0.685	0.618	0.808
K	0.014	0.013	0.014	0.017	0.023	0.021	0.024	0.021	0.010	0.021
An	41.7	44.1	38.3	31.7	29.9	29.3	29.1	30.3	39.4	18.1
Ab	56.9	54.6	60.3	66.6	68.0	68.7	68.6	67.7	59.7	79.8
Or	1.4	1.3	1.4	1.7	2.2	2.0	2.4	2.0	0.9	2.1

n = one spot analysis of “n” different crystals; Mega core = megacrystic core; Mega rim = megacrystic rim. wt% values are the average of n analyses. An, Ab, and Or are the feldspar end-members expressed as mole% that sum to 100%. FeO^t, total Fe as FeO.

and Table 2 and Supplemental Tables S1–S3. Whole-rock major and trace element (Sr, Ba, Zr, Y, Sc, Cr, and Ni) analysis of the megacrystic amphibolite matrix was performed at Montclair State University by inductively coupled plasma optical emission spectrometry on a JY Ultima C system. Additional trace element (REE, Co, V, Nb, Ta, Hf, U, Th, Rb, and Cs) analysis was performed at Binghamton University by inductively coupled plasma mass spectrometry using a Perkin-Elmer Elan 6000. Details of the analytical methods can be found in [29]. All whole-rock geochemical data are presented in Table 3 and Table 4. A single plagioclase megacryst from sample BLR205 was analyzed for oxygen isotope ratios. About 2 mg of pure plagioclase was hand-picked under a binocular microscope and fluorinated with BrF₅ while being heated using a CO₂ laser. The evolved oxygen was converted to CO₂ for mass spectrometry using a Finnigan MAT 251 mass spectrometer (Bremen, Germany) following the methods of [31]. This analysis was not duplicated, but reproducibility of standards and replicate samples during the analytical session was ±0.6‰ (2σ).

4. Megacrystic Amphibolites

4.1. Field Relationships

Megacrystic amphibolites are recognized from five locations that span the western and eastern New Jersey Highlands (Figure 2). They form dike-like bodies 0.5 to 2 m wide that have sharp, conformable contacts against adjacent country rocks and contain no xenoliths (Figure 3). The length of amphibolite bodies is indeterminate because they disappear beneath a veneer of overburden along strike; however, outcrops sampled for this study ranged from 10 to 50 m in length along strike. The concordance of contacts argues for the tectonic transposition of the amphibolites and their emplacement prior to the onset of Ottawan high-grade metamorphism. Two of the amphibolites (EAS2 and TQ423) intrude tonalite gneiss of the Losee Suite, two amphibolites (ST106, DOV189) intrude hedenbergite granite, and one amphibolite (BLR205) intrudes hastingsite granite. The occurrence of megacrystic amphibolites in an east–west array (Figure 2), and the lack of a common lithologic association, are consistent with their emplacement as dikes. Sample EAS2 falls off of and well south of the main trend, suggesting that it may be part of a separate parallel array of megacrystic amphibolites that extend to the east beneath the Mesozoic Piedmont (Figure 2). All of the amphibolites display a penetrative high-strain metamorphic foliation defined by the alignment of amphibole and biotite.

The apparent absence of minerals suitable for U-Pb geochronology such as zircon or monazite precludes the ability to directly date the amphibolites, although reasonable inferences on their age may be made indirectly from the field relationships. Intrusion of the amphibolites into hastingsite and hedenbergite granites provides a lower bounding age of 1182 Ma, and the presence of conformable contacts and a high-strain metamorphic fabric provides an upper bounding age for the amphibolites of ca. 1024 Ma. Despite the acquisition of a considerable amount of U-Pb zircon geochronology in the study area, rocks are not recognized in the age range of ca. 1080–1050 Ma. We thus show in later sections that our results are more consistent with an age of 1160–1130 Ma for the megacrystic amphibolites.

4.2. Mineralogy

The amphibolites in this study contain plagioclase megacrysts typically <8 cm long (Figure 3), but locally are as long as 13 cm, in a medium-grained matrix composed of plagioclase, amphibole, biotite, clinopyroxene, Fe-Ti oxides, and apatite. Plagioclase megacrysts appear to be evenly distributed within single outcrops and they lack embayed contacts with other matrix minerals or evidence for resorption (Figure 3). Plagioclase and amphibole are the most abundant minerals in the matrix that together with biotite and Fe-Ti oxides are found in all of the samples. Clinopyroxene is present in minor amounts in samples TQ423 and BLR205, and apatite occurs in trace amounts in samples EAS2 and BLR205.

Representative analyses of core spot and rim spot compositions of plagioclase megacrysts and matrix plagioclase are given in Table 1. Because of their unusually large size, megacrystic plagioclases presented a challenge to analyze and we present what is considered to be the most representative analyses of cores and rims. Matrix plagioclase comprises most of the groundmass in amphibolites. It has the composition of andesine (An_31–39), with some samples from DOV189 that are less calcic, having the composition of oligoclase (An_13–18). The anorthite composition in megacryst rims in BLR205 (An_27–32) and EAS2 (An_29–31) matches the matrix plagioclase (BLR205 = An_29–30, EAS2 = An_28–31), indicating they are in equilibrium, whereas the megacryst rim in TQ 423-3 has higher anorthite (An_44–45) than the core (An_40–43) or the matrix (An_36–41), suggesting conditions of disequilibrium. Similar comparisons could not be made for ST106 or DOV189, but we note that matrix plagioclase in ST106 (An_37–39) is comparable to matrix plagioclase in TQ423-3 in contrast to DOV189 (An_17–21) that is more sodic than all other samples. Concentrations of K₂O are low in all megacrystic and matrix plagioclase that have Or_0.9–2.4, consistent with the observed absence of perthitic lamellae or of sericitic alteration.

Representative analyses of amphiboles and their structural formula are given in Table 2. Based on the amphibole classification of [32], they belong to the calcic amphibole group. The crystals analyzed are magnesio-hastingsite that have <7.00 atoms of Si per formula unit (apfu), <0.50 apfu Ti, >1.50 apfu Ca, and >0.50 apfu Na + K. All of the amphiboles are Mg-rich and have X_Mg (Mg/Mg + Fe²⁺) that ranges from 0.42 to 0.64, and they have high contents of TiO₂ (1.41–2.67 wt.%) and FeO (16.07–22.38 wt.%). A comparison of core and rim compositions reveals very minor compositional variations, suggesting the absence of zoning in any of the crystals analyzed.

Table 2. Representative EMPA of amphibole.

Sample	TQ423-3			TQ423-4			EAS2	BLR205	ST106	DOV189
Note	Core	Rim	Matrix	Core	Rim	Matrix	Matrix	Matrix	Matrix	Core	Rim	Matrix
	n = 5	n = 3	n = 21	n = 4	n = 4	n = 7	n = 42	n = 20	n = 18	n = 5	n = 4	n = 8
Oxides in wt.%
SiO2	42.27	41.90	42.71	42.23	41.30	42.14	42.31	41.58	38.17	41.26	41.45	41.50
TiO2	2.08	1.91	1.97	2.08	1.93	2.02	2.06	2.30	1.41	2.39	2.63	2.67
Al2O3	12.92	12.61	12.22	11.65	11.68	11.70	11.22	11.76	14.12	11.03	11.10	10.98
FeOt	16.54	16.52	16.07	19.95	20.15	20.00	17.98	19.12	19.13	22.38	22.07	22.05
MnO	0.30	0.29	0.28	0.31	0.32	0.31	0.34	0.25	0.62	0.30	0.32	0.32
MgO	10.47	10.56	11.25	8.73	8.82	8.75	10.04	8.57	7.16	7.48	7.36	7.47
CaO	10.96	10.95	11.08	11.14	11.10	11.16	11.46	11.14	10.70	10.59	10.67	10.66
Na2O	2.10	2.08	2.11	1.60	1.60	1.59	1.73	1.69	1.97	2.28	2.27	2.31
K2O	1.00	1.02	0.99	1.59	1.58	1.60	1.87	1.85	1.94	1.78	1.78	1.74
Total	98.64	97.84	98.67	99.28	98.48	99.26	99.02	98.25	95.23	99.49	99.65	99.69
Structural formulae based on 23 oxygens, 15 cations
Si	6.191	6.190	6.236	6.275	6.186	6.262	6.295	6.275	5.973	6.224	6.250	6.252
AlIV	1.809	1.810	1.764	1.725	1.814	1.738	1.705	1.725	2.027	1.776	1.750	1.748
AlVI	0.422	0.385	0.338	0.315	0.248	0.311	0.263	0.366	0.578	0.185	0.223	0.201
Ti	0.229	0.212	0.216	0.232	0.217	0.226	0.231	0.261	0.166	0.271	0.298	0.303
Fe3+	0.706	0.747	0.745	0.635	0.801	0.660	0.474	0.384	0.544	0.593	0.476	0.492
Fe2+	1.320	1.294	1.217	1.844	1.723	1.826	1.764	2.029	1.960	2.230	2.307	2.286
Mn	0.037	0.036	0.035	0.039	0.041	0.039	0.043	0.032	0.082	0.038	0.041	0.041
Mg	2.286	2.326	2.449	1.934	1.970	1.938	2.227	1.928	1.670	1.682	1.655	1.678
Ca	1.720	1.733	1.733	1.774	1.781	1.777	1.827	1.801	1.794	1.725	1.724	1.721
Na	0.596	0.596	0.597	0.461	0.465	0.458	0.499	0.494	0.598	0.664	0.664	0.674
K	0.187	0.192	0.184	0.301	0.302	0.303	0.355	0.356	0.387	0.343	0.342	0.334
Mg/Mg + Fe	0.634	0.642	0.668	0.512	0.533	0.515	0.558	0.487	0.460	0.430	0.418	0.423
AlT	2.231	2.195	2.102	2.040	2.062	2.049	1.968	2.091	2.605	1.961	1.973	1.949
P (MPa)	598	582	543	517	526	521	486	538	756	484	488	478
T °C (avg.)	750			767			761	723	756	765

n = one spot analysis of “n” different crystals. Al^T = Al^IV + Al^VI. T calculated from hbd-plag pairs using geothermometer of [33]. P calculated from geobarometer of [34]. FeO^t, total Fe as FeO.

Table 2. Representative EMPA of amphibole.

Sample	TQ423-3			TQ423-4			EAS2	BLR205	ST106	DOV189
Note	Core	Rim	Matrix	Core	Rim	Matrix	Matrix	Matrix	Matrix	Core	Rim	Matrix
	n = 5	n = 3	n = 21	n = 4	n = 4	n = 7	n = 42	n = 20	n = 18	n = 5	n = 4	n = 8
Oxides in wt.%
SiO2	42.27	41.90	42.71	42.23	41.30	42.14	42.31	41.58	38.17	41.26	41.45	41.50
TiO2	2.08	1.91	1.97	2.08	1.93	2.02	2.06	2.30	1.41	2.39	2.63	2.67
Al2O3	12.92	12.61	12.22	11.65	11.68	11.70	11.22	11.76	14.12	11.03	11.10	10.98
FeOt	16.54	16.52	16.07	19.95	20.15	20.00	17.98	19.12	19.13	22.38	22.07	22.05
MnO	0.30	0.29	0.28	0.31	0.32	0.31	0.34	0.25	0.62	0.30	0.32	0.32
MgO	10.47	10.56	11.25	8.73	8.82	8.75	10.04	8.57	7.16	7.48	7.36	7.47
CaO	10.96	10.95	11.08	11.14	11.10	11.16	11.46	11.14	10.70	10.59	10.67	10.66
Na2O	2.10	2.08	2.11	1.60	1.60	1.59	1.73	1.69	1.97	2.28	2.27	2.31
K2O	1.00	1.02	0.99	1.59	1.58	1.60	1.87	1.85	1.94	1.78	1.78	1.74
Total	98.64	97.84	98.67	99.28	98.48	99.26	99.02	98.25	95.23	99.49	99.65	99.69
Structural formulae based on 23 oxygens, 15 cations
Si	6.191	6.190	6.236	6.275	6.186	6.262	6.295	6.275	5.973	6.224	6.250	6.252
AlIV	1.809	1.810	1.764	1.725	1.814	1.738	1.705	1.725	2.027	1.776	1.750	1.748
AlVI	0.422	0.385	0.338	0.315	0.248	0.311	0.263	0.366	0.578	0.185	0.223	0.201
Ti	0.229	0.212	0.216	0.232	0.217	0.226	0.231	0.261	0.166	0.271	0.298	0.303
Fe3+	0.706	0.747	0.745	0.635	0.801	0.660	0.474	0.384	0.544	0.593	0.476	0.492
Fe2+	1.320	1.294	1.217	1.844	1.723	1.826	1.764	2.029	1.960	2.230	2.307	2.286
Mn	0.037	0.036	0.035	0.039	0.041	0.039	0.043	0.032	0.082	0.038	0.041	0.041
Mg	2.286	2.326	2.449	1.934	1.970	1.938	2.227	1.928	1.670	1.682	1.655	1.678
Ca	1.720	1.733	1.733	1.774	1.781	1.777	1.827	1.801	1.794	1.725	1.724	1.721
Na	0.596	0.596	0.597	0.461	0.465	0.458	0.499	0.494	0.598	0.664	0.664	0.674
K	0.187	0.192	0.184	0.301	0.302	0.303	0.355	0.356	0.387	0.343	0.342	0.334
Mg/Mg + Fe	0.634	0.642	0.668	0.512	0.533	0.515	0.558	0.487	0.460	0.430	0.418	0.423
AlT	2.231	2.195	2.102	2.040	2.062	2.049	1.968	2.091	2.605	1.961	1.973	1.949
P (MPa)	598	582	543	517	526	521	486	538	756	484	488	478
T °C (avg.)	750			767			761	723	756	765

n = one spot analysis of “n” different crystals. Al^T = Al^IV + Al^VI. T calculated from hbd-plag pairs using geothermometer of [33]. P calculated from geobarometer of [34]. FeO^t, total Fe as FeO.

Biotite is subordinate in abundance to amphibole and forms small disseminated crystals, most of which are oriented parallel to or at low angles to amphibole. All of the crystals sampled have compositions characterized by X_Mg of 0.40–0.66 and high FeO (16.10–21.66 wt.%; these values represent total ferrous iron obtained directly from the EMPA; Table S1), similar to values in amphibole. However, concentrations of TiO₂ are somewhat higher (3.24–5.17 wt.%) than in amphibole. Comparable contents of high TiO₂ in both amphibole and biotite indicate overlapping crystallization of these phases under conditions of high temperature. This will be discussed in more detail in a later section.

Clinopyroxene forms small minor crystals in the matrix that are associated with amphibole. It is calcium-rich (0.86–0.94 apfu Ca) and has the composition of diopside with En_33–37, Fs_17–22, and Wo_45–48 (Table S2). Fe-Ti oxides consist of ilmenite and magnetite that form small, subangular crystals disseminated throughout the rock matrix and less commonly occur as intergrowths of ilmenite and magnetite. Representative analyses of Fe-Ti oxides are given in Table S3. Ilmenite has a relatively uniform composition in contrast to magnetite, which ranges from low-Ti magnetite (0.06–1.10 wt.% TiO₂) to titanomagnetite that contains 4.83–27.37 wt.% TiO₂. Apatite was not recognized in all samples, and where present it occurs as small crystals aligned mainly along amphibole crystal boundaries. Apatite content is highly variable among the amphibolites sampled, with EAS2 and DOV189 containing as much as 0.95 wt.% P₂O₅ and 1.5–2.2 normative % apatite.

4.3. Whole-Rock Geochemistry

Whole-rock major and trace element data of the megacrystic amphibolites (matrix only) in this study are given in Table 3. Major element analysis of the amphibolite matrix yielded the following wt.% ranges (SiO₂ = 46%–51%, TiO₂ = 0.7%–3.3%, Al₂O₃ = 14%–23%, Fe₂O₃^t = 7%–16%, MgO = 2%–6%, CaO = 7%–10%, Na₂O = 3%–5%, and K₂O = 1%–2.4%). Mg# and compatible trace element abundances are generally low, but span a broad range (Mg# = 40–51, Cr = 18–89 ppm, Ni = 18–64 ppm, Co = 10–56 ppm). Key major element characteristics of the megacrystic amphibolite in this study are shown in Figure 4. The matrix of megacrystic amphibolites has whole-rock major element abundances that straddle the boundary between alkalic and tholeiitic compositions (Figure 4a) and specifically has a strong affinity to high-Fe-Al tholeiitic magma types (Figure 4b). Megacrystic amphibolites also have comparable major element abundances to AMCG-related mafic rocks from the Adirondack Highlands [13,14] and mafic dikes associated with anorthosite from Tanzania [8].

Table 3. Whole-rock composition of megacrystic amphibolites in the New Jersey Highlands compared to anorthosite-related mafic rocks from the Adirondacks, NY, and Tanzania.

Sample	TQ423-3	TQ423-4	EAS2	BLR205	ST106	DOV189	P4	Adir	A161-162
Lithology	MA	MA	MA	MA	MA	MA	OAG	AM	AM
Oxides in wt.%
SiO2	46.22	46.98	45.08	46.33	50.90	46.58	44.10	48.27	47.48
TiO2	1.76	2.00	2.56	2.09	0.70	3.31	1.31	2.66	1.66
Al2O3	17.46	16.64	15.45	15.62	23.01	14.02	14.70	15.84	16.97
Fe2O3t	12.93	13.77	14.20	13.30	6.72	16.11	15.50	13.37	12.39
MnO	0.18	0.16	0.25	0.18	0.16	0.23	0.21	0.20	0.17
MgO	5.83	5.62	5.85	5.89	2.04	4.64	8.42	5.89	7.95
CaO	9.36	7.25	8.23	9.87	8.98	8.50	11.60	8.75	9.80
Na2O	3.48	3.26	3.02	3.46	4.82	4.27	1.46	3.29	2.42
K2O	0.98	2.43	2.28	1.27	1.20	1.19	1.50	1.29	0.36
P2O5	0.24	0.26	0.95	0.55	0.08	0.67	0.17	0.45	0.16
LOI	1.05	1.31	1.30	0.55	2.05	n.d.	0.90	n.d.	n.d.
Total	99.49	99.69	99.19	99.11	100.66	99.52	99.87	100.01	99.66
Mg#	51.2	48.7	49.0	50.7	41.5	40.1	51.8	46.1	57.5
Trace elements in ppm
Rb	23	73	70	24	46	27	31	32	<10
Ba	301	684	484	169	455	318	266	286	261
Sr	422	351	367	322	588	349	208	277	291
Cs	0.49	1.38	0.67	0.14	0.57	0.08	n.d.	n.d.	<10
Zr	122	145	215	233	45	237	n.d.	204	102
Y	29.2	32.2	42.8	55.1	22.1	44.7	n.d.	n.d.	15
Nb	3.4	4.0	9.8	18.7	2.2	13.5	n.d.	n.d.	<10
Hf	3.32	3.73	4.56	5.23	1.17	5.67	0.89	n.d.	2.70
Ta	0.20	0.23	0.60	1.90	0.12	0.82	0.20	n.d.	<5.0
Th	0.55	0.70	0.88	5.20	1.71	3.76	0.14	n.d.	0.46
U	0.37	0.72	0.49	3.25	0.95	2.12	n.d.	n.d.	0.19
Cr	42	36	54	45	18	69	200	n.d.	84
Ni	55	64	64	51	18	37	n.d.	55	108
Co	50	56	47	40	10	43	34	n.d.	139
Sc	23.4	25.0	24.8	27.8	13.5	34.7	40.9	30	31.5
V	209	221	236	237	94	311	n.d.	n.d.	252
La	7.3	10.9	26.2	26.7	30.7	37.9	13.3	24.5	6.2
Ce	19.7	24.5	55.5	62.5	40.1	69.8	24.2	64.2	15.4
Pr	3.1	3.6	8.3	8.9	4.2	8.9	n.d.	n.d.	2.4
Nd	15.1	16.6	37.1	38.0	14.8	37.3	5.7	38.0	11.4
Sm	4.60	4.77	9.26	9.78	3.27	9.55	3.61	9.22	3.37
Eu	1.54	1.90	2.69	2.13	1.41	2.91	1.50	2.66	1.42
Gd	4.67	5.24	8.36	8.95	3.71	8.69	n.d.	n.d.	3.85
Tb	0.85	0.93	1.41	1.54	0.63	1.45	0.85	1.56	0.65
Dy	5.12	5.71	7.58	8.91	3.66	7.95	n.d.	n.d.	3.82
Ho	1.13	1.17	1.68	2.06	0.76	1.79	n.d.	n.d.	0.77
Er	2.93	3.21	4.20	5.25	1.97	4.22	n.d.	n.d.	2.15
Tm	0.46	0.53	0.62	0.89	0.32	0.67	n.d.	n.d.	0.32
Yb	2.78	3.00	3.61	5.53	1.89	4.04	3.65	5.12	1.95
Lu	0.41	0.46	0.57	0.86	0.31	0.59	0.55	0.80	0.30
Eu/Eu*	1.03	1.17	0.95	0.70	1.25	0.99	1.15	n.d.	1.21

Whole-rock composition of megacrystic amphibolite (MA) TQ423-3 to DOV189 [this study] and gabbroic anorthosite (GA) P4 beneath anorthosite sheet, New Jersey Highlands [23]. Adir, average of 20 amphibolites (AM) associated with anorthosite, Adirondack Highlands [14]. A161-162, average of two mafic dikes associated with anorthosite, Tanzania [8]. Fe₂O₃^t, total Fe as Fe₂O₃; LOI, loss on ignition; n.d., not determined.

Table 3. Whole-rock composition of megacrystic amphibolites in the New Jersey Highlands compared to anorthosite-related mafic rocks from the Adirondacks, NY, and Tanzania.

Sample	TQ423-3	TQ423-4	EAS2	BLR205	ST106	DOV189	P4	Adir	A161-162
Lithology	MA	MA	MA	MA	MA	MA	OAG	AM	AM
Oxides in wt.%
SiO2	46.22	46.98	45.08	46.33	50.90	46.58	44.10	48.27	47.48
TiO2	1.76	2.00	2.56	2.09	0.70	3.31	1.31	2.66	1.66
Al2O3	17.46	16.64	15.45	15.62	23.01	14.02	14.70	15.84	16.97
Fe2O3t	12.93	13.77	14.20	13.30	6.72	16.11	15.50	13.37	12.39
MnO	0.18	0.16	0.25	0.18	0.16	0.23	0.21	0.20	0.17
MgO	5.83	5.62	5.85	5.89	2.04	4.64	8.42	5.89	7.95
CaO	9.36	7.25	8.23	9.87	8.98	8.50	11.60	8.75	9.80
Na2O	3.48	3.26	3.02	3.46	4.82	4.27	1.46	3.29	2.42
K2O	0.98	2.43	2.28	1.27	1.20	1.19	1.50	1.29	0.36
P2O5	0.24	0.26	0.95	0.55	0.08	0.67	0.17	0.45	0.16
LOI	1.05	1.31	1.30	0.55	2.05	n.d.	0.90	n.d.	n.d.
Total	99.49	99.69	99.19	99.11	100.66	99.52	99.87	100.01	99.66
Mg#	51.2	48.7	49.0	50.7	41.5	40.1	51.8	46.1	57.5
Trace elements in ppm
Rb	23	73	70	24	46	27	31	32	<10
Ba	301	684	484	169	455	318	266	286	261
Sr	422	351	367	322	588	349	208	277	291
Cs	0.49	1.38	0.67	0.14	0.57	0.08	n.d.	n.d.	<10
Zr	122	145	215	233	45	237	n.d.	204	102
Y	29.2	32.2	42.8	55.1	22.1	44.7	n.d.	n.d.	15
Nb	3.4	4.0	9.8	18.7	2.2	13.5	n.d.	n.d.	<10
Hf	3.32	3.73	4.56	5.23	1.17	5.67	0.89	n.d.	2.70
Ta	0.20	0.23	0.60	1.90	0.12	0.82	0.20	n.d.	<5.0
Th	0.55	0.70	0.88	5.20	1.71	3.76	0.14	n.d.	0.46
U	0.37	0.72	0.49	3.25	0.95	2.12	n.d.	n.d.	0.19
Cr	42	36	54	45	18	69	200	n.d.	84
Ni	55	64	64	51	18	37	n.d.	55	108
Co	50	56	47	40	10	43	34	n.d.	139
Sc	23.4	25.0	24.8	27.8	13.5	34.7	40.9	30	31.5
V	209	221	236	237	94	311	n.d.	n.d.	252
La	7.3	10.9	26.2	26.7	30.7	37.9	13.3	24.5	6.2
Ce	19.7	24.5	55.5	62.5	40.1	69.8	24.2	64.2	15.4
Pr	3.1	3.6	8.3	8.9	4.2	8.9	n.d.	n.d.	2.4
Nd	15.1	16.6	37.1	38.0	14.8	37.3	5.7	38.0	11.4
Sm	4.60	4.77	9.26	9.78	3.27	9.55	3.61	9.22	3.37
Eu	1.54	1.90	2.69	2.13	1.41	2.91	1.50	2.66	1.42
Gd	4.67	5.24	8.36	8.95	3.71	8.69	n.d.	n.d.	3.85
Tb	0.85	0.93	1.41	1.54	0.63	1.45	0.85	1.56	0.65
Dy	5.12	5.71	7.58	8.91	3.66	7.95	n.d.	n.d.	3.82
Ho	1.13	1.17	1.68	2.06	0.76	1.79	n.d.	n.d.	0.77
Er	2.93	3.21	4.20	5.25	1.97	4.22	n.d.	n.d.	2.15
Tm	0.46	0.53	0.62	0.89	0.32	0.67	n.d.	n.d.	0.32
Yb	2.78	3.00	3.61	5.53	1.89	4.04	3.65	5.12	1.95
Lu	0.41	0.46	0.57	0.86	0.31	0.59	0.55	0.80	0.30
Eu/Eu*	1.03	1.17	0.95	0.70	1.25	0.99	1.15	n.d.	1.21

Whole-rock composition of megacrystic amphibolite (MA) TQ423-3 to DOV189 [this study] and gabbroic anorthosite (GA) P4 beneath anorthosite sheet, New Jersey Highlands [23]. Adir, average of 20 amphibolites (AM) associated with anorthosite, Adirondack Highlands [14]. A161-162, average of two mafic dikes associated with anorthosite, Tanzania [8]. Fe₂O₃^t, total Fe as Fe₂O₃; LOI, loss on ignition; n.d., not determined.

Key whole-rock trace element characteristics of megacrystic amphibolites (matrix only) are shown in Figure 5. They have broadly similar chondrite-normalized REE patterns (Figure 5a) that display enrichment in LREEs (La/Yb)_N = 1.73–10.69) and relatively flat MREEs and HREEs (Gd/Yb)_N = 1.30–1.85, and except for sample ST106, they have small positive Eu anomalies (Eu/Eu* = 0.95–1.25). On a mid-ocean ridge basalt (MORB)-normalized multi-element diagram (Figure 5b), megacrystic amphibolites display patterns that are enriched in LILEs and depleted in HFSEs and thus result in high LILE/HFSE ratios. The Ba/Nb ratio in all samples except for BLR205 ranges from 24 to 211, comparable to values from subduction-related mafic magmas from the Andean Southern Volcanic Zone [37]. As with the major elements, the trace element concentrations in our megacrystic amphibolites are comparable to AMCG-related mafic rocks from the Adirondack Highlands [13,14].

4.4. Oxygen Isotope Data

The oxygen isotope ratio measured for plagioclase from sample BLR205 has a δ¹⁸O value of 6.45‰ (VSMOW) that is consistent with a magmatic whole-rock δ¹⁸O of 6.2‰ (using fractionations from [38]). This is comparable to metagabbro related to AMCG suite plutonism in the Adirondack Highlands (6.8 ± 0.3‰, n = 16; [39,40]). Mantle melts typically have oxygen isotope ratios in equilibrium with plagioclase δ¹⁸O of 5.6 to 6.0‰ [38], whereas the values of BLR205 are slightly higher and comparable to mafic rocks that have experienced limited crustal interaction or minor modification of the source by previous slab-derived melt or subducted sediment. We return to a discussion of the effect of these processes on the megacrystic amphibolites later in this paper. However, contamination by more than a nominal amount of assimilant is not likely, as anorthosites and related rocks elsewhere that have a large component of supracrustal input show δ¹⁸O values of 8–10‰ [40]. No more than 10–20% contamination of a mantle melt by upper-crustal lithologies (e.g., [41]) would be consistent with the δ¹⁸O of 6.45‰ measured from plagioclase in BLR205.

5. Discussion

5.1. Crystallization Conditions

The initial crystallization temperature of the megacrystic amphibolites is uncertain because of their geochemically evolved composition and their subsequent modification under granulite facies metamorphic conditions. Estimates of the temperature during the metamorphic peak were made using the amphibole–plagioclase geothermometer of [33] from multiple pairs of adjacent minerals for each sample. The thermometer was calibrated for calcic amphiboles and is therefore appropriate for use with our samples. Calculated average temperatures are 755 °C (n = 8) for TQ423, 767 °C (n = 3) for EAS2, 723 °C (n = 6) for BLR205, 756 °C (n = 5) for ST106, and 765 °C (n = 3) for DOV189. The Ti-in-biotite geothermometer [42] provides temperature estimates with a precision of ±12 °C at the upper end of its calibration range of 480–800 °C for biotite having X_Mg = 0.275–1.00 and Ti = 0.04–0.60 apfu in pelitic rocks with coexisting ilmenite and rutile that equilibrated at a pressure range of 400–600 MPa. Although the samples in this study lack rutile and are not pelites (e.g., garnet, muscovite, and sillimanite are not present), they contain ilmenite and have biotite with X_Mg that ranges from 0.421 to 0.592 and Ti ranging from 0.478 to 0.660 apfu, and are therefore appropriate for use with this thermometer. Where possible, we used biotites that were adjacent to amphibole–plagioclase pairs for which geothermometry was used, but we note that the results from discrete crystals of biotite define a narrow range. The Ti-in-biotite geothermometer yielded temperature estimates that range from 731 to 786 °C and are in good agreement with the results using amphibole and plagioclase.

The primary igneous assemblage of the megacrystic amphibolites was composed of plagioclase, clinopyroxene, Fe-Ti oxides, and apatite that subsequently equilibrated at granulite facies metamorphic conditions. The abundant amphibole, biotite, and magnetite formed under amphibolite facies metamorphic conditions at some time following emplacement of the amphibolites, most likely during Ottawan orogenesis. We propose that the comparable temperature estimates for all samples imply they formed under similar magmatic conditions, and that variations in the mineralogy are due to modification by post-emplacement processes. This may explain the diopsidic composition of clinopyroxene in our samples rather than augite in that metamorphic recrystallization may result in more calcic pyroxene [12].

Estimates of the confining pressure were made using the Al-in-hornblende geobarometer of [34] that is calibrated for pressures of 200–800 MPa and temperatures of 720–780 °C. The geobarometer uses Al^t (the total of Al^IV and Al^VI) from an amphibole structural formula calculated from 23 oxygens and 15 cations because experimental studies have shown a correlation between increasing pressure and higher total Al. Calculated pressures range from 517 to 598 MPa for TQ423 and BLR205 and from 484 to 488 MPa for EAS2 and DOV189. Amphiboles from ST106 yielded unusually high estimates of 756–761 MPa. For the most part, the pressure and temperature estimates in this study are comparable to those of previous studies in the New Jersey Highlands and support an interpretation of metamorphism of the amphibolites under peak conditions of low-pressure and high-temperature granulite facies.

5.2. Origin of Plagioclase Megacrysts

The simplest interpretation to account for the accumulation of plagioclase megacrysts in the amphibolites sampled is through density separation and flotation. Experimental studies show that at crustal pressures, plagioclase in Fe-rich tholeiitic magmas having <50 wt.% SiO₂ and 12.5–15 wt.% FeO (comparable to our amphibolite samples) would be positively buoyant and float to the roof of the magma chamber (e.g., [43]). Plagioclase megacrysts in our study have a calculated density of 2.66–2.69 g/cm³ that contrasts with the whole-rock density of the host amphibolites (2.81–3.04 g/cm³) and gabbro from the study area (2.94–3.02 g/cm³), with the latter presumed to approximate that of the mafic source. The density contrast is thus permissive of separation of the megacrysts from the melt by flotation. Furthermore, the Fe-rich composition of the amphibolites and the presence of an oxide–apatite gabbronorite layer stratigraphically beneath anorthosite in the study area support the interpretation of separation of the magma into a denser cumulate layer through settling in a manner reported by [12] for similar lithologies in the Adirondack Highlands.

An alternative interpretation to flotation for the accumulation of plagioclase megacrysts involves some combination of mixing of megacryst-rich material and mafic magma in a periodically recharging magma chamber (e.g., [43,44,45,46,47]). As noted above, the anorthite values of megacrystic plagioclase rims in BLR205 and EAS2 are in equilibrium with matrix plagioclase, whereas the megacryst rim in TQ423-3 has higher anorthite than the core or matrix plagioclase, suggesting conditions of disequilibrium. The observed compositional differences between megacrystic plagioclase in the amphibolites of our study may be due to the introduction of more than one pulse of recharging magma into the chamber that would account for the more calcic rims in some megacrysts that formed in response to an increase in the temperature of the magma. Conversely, experimental studies show that increasing pressure during plagioclase crystallization results in an increase in the Ab content [48]. In that case, compositional differences may have resulted from variation in pressure during ascent and emplacement of the magma.

Our preferred model for the origin and accumulation of megacrystic plagioclase incorporates elements of both magma mixing and flotation. We envisage a sequence of events that consisted of the extraction of a plagioclase-rich crystal-liquid mush from partially formed anorthosite that mixed in the magma chamber with the mafic source of the amphibolites. Given the volume, size, and composition of the plagioclase megacrysts, the most likely source is anorthosite. Density separation resulted in the accumulation of megacrystic plagioclase in the upper part of the magma chamber. In the New Jersey Highlands, the ascent of this mixed megacryst-rich mafic melt along an array of developing extensional fractures led to their localized emplacement as dikes. This extensional tectonic setting may have interrupted the more extensive fractionation and plagioclase separation necessary to form voluminous anorthosite intrusions.

5.3. Correlation with Regional Mafic Intrusions and Anorthositic Magmatism

Given the volume, size, and composition of the plagioclase megacrysts, and the Al-Fe composition of the amphibolites, a link to anorthosite is consistent with the available data. Megacrystic amphibolites in the New Jersey Highlands are lithologically and geochemically similar to mafic rocks associated with anorthosite massifs in the Adirondack Highlands described by [12,14]. Both the New Jersey and Adirondack mafic rocks are tholeiitic (Y/Nb = 4.7 and 5.3, respectively) olivine-normative basalts. They have similar geochemical characteristics, including high TiO₂ (2.5 to 3.5 wt.%), Al₂O₃ (14 to 19 wt.%), and Fe₂O₃^t (10 to 16 wt.%) (Table 3), and comparable trace and REE abundances that form a distinct high-Al-Fe mafic magma type characteristic of anorthosite associations [13]. Additionally, as noted above, mafic intrusions and anorthosite in the Adirondack Highlands [12,15] and Grenville Province [49] are spatially and temporally associated with thin layers of oxide- and apatite-rich gabbronorite that are similar mineralogically and geochemically to the thin oxide- and apatite-rich layer structurally beneath gabbroic anorthosite and anorthosite reported by [23] in the New Jersey Highlands (Table 4, Sample P6). Although undated, anorthosites in the New Jersey Highlands and Honey Brook Upland of Pennsylvania (Figure 1) have compositions that are comparable to ca. 1155 Ma anorthosite in the Adirondack Highlands, while alkalic anorthosite containing highly exolved antiperthitic plagioclase in Grenvillian inliers in the southern Appalachians yielded much younger ages of 1040–1010 Ma [20,21].

Table 4. Whole-rock composition of mafic rocks associated with anorthosite from the New Jersey Highlands and New York Hudson Highlands.

Sample	PP43	ST114	TQ9a	TQ9b	P4	P6	SF10	SF11
Lithology	G	G	G	G	GA	OAG	AM	AM
Location	NJH	NJH	NJH	NJH	NJH	NJH	NYHH	NYHH
Oxides in wt.%
SiO2	46.70	50.77	48.63	49.87	44.10	38.60	48.40	49.22
TiO2	0.51	1.08	2.49	1.99	1.31	7.19	1.14	1.60
Al2O3	21.00	14.26	14.91	16.41	14.70	12.20	18.39	15.65
Fe2O3t	9.28	12.06	13.94	11.73	15.50	20.67	9.99	13.93
MnO	0.14	0.18	0.19	0.16	0.21	0.32	0.22	0.25
MgO	6.46	6.67	4.64	3.93	8.42	5.67	6.62	5.64
CaO	13.50	10.08	9.06	8.49	11.60	10.40	9.69	8.61
Na2O	1.60	3.40	4.20	4.66	1.46	1.33	4.16	4.15
K2O	0.39	0.38	1.19	1.47	1.50	0.92	1.15	0.64
P2O5	0.08	0.13	0.25	0.24	0.17	2.06	0.24	0.31
LOI	0.23	0.68	n.d.	n.d.	0.90	1.00	n.d.	n.d.
Total	99.89	99.69	99.49	98.96	99.87	100.36	100.0	100.0
Mg#	57.9	49.6	43.7	43.9	51.8	n.d.	56.0	43.9
Trace elements in ppm
Ba	60	53	372	581	266	724	n.d.	n.d.
Sr	810	180	457	569	208	n.d.	n.d.	n.d.
Zr	n.d.	34	293	187	n.d.	335	n.d.	n.d.
Y	n.d.	n.d.	67.6	31	n.d.	n.d.	n.d.	n.d.
Cr	40	n.d.	22	15	200	36.4	n.d.	n.d.
Ni	n.d.	n.d.	25	23	n.d.	n.d.	n.d.	n.d.
Sc	n.d.	n.d.	32	25.2	41	58.6	n.d.	n.d.

Whole-rock composition of gabbro (G) PP43, ST114 from [35] and TQ9a, TQ9b [this study]; gabbroic anorthosite (GA) P4 and oxide–apatite gabbro (OAG) P6 beneath anorthosite sheet from [23]; and amphibolite (AM) SF10, SF11 associated with anorthosite [this study]. NJH, New Jersey Highlands; NYHH, New York Hudson Highlands. Fe₂O₃^t, total Fe as Fe₂O₃; LOI, loss on ignition; n.d., not determined.

Table 4. Whole-rock composition of mafic rocks associated with anorthosite from the New Jersey Highlands and New York Hudson Highlands.

Sample	PP43	ST114	TQ9a	TQ9b	P4	P6	SF10	SF11
Lithology	G	G	G	G	GA	OAG	AM	AM
Location	NJH	NJH	NJH	NJH	NJH	NJH	NYHH	NYHH
Oxides in wt.%
SiO2	46.70	50.77	48.63	49.87	44.10	38.60	48.40	49.22
TiO2	0.51	1.08	2.49	1.99	1.31	7.19	1.14	1.60
Al2O3	21.00	14.26	14.91	16.41	14.70	12.20	18.39	15.65
Fe2O3t	9.28	12.06	13.94	11.73	15.50	20.67	9.99	13.93
MnO	0.14	0.18	0.19	0.16	0.21	0.32	0.22	0.25
MgO	6.46	6.67	4.64	3.93	8.42	5.67	6.62	5.64
CaO	13.50	10.08	9.06	8.49	11.60	10.40	9.69	8.61
Na2O	1.60	3.40	4.20	4.66	1.46	1.33	4.16	4.15
K2O	0.39	0.38	1.19	1.47	1.50	0.92	1.15	0.64
P2O5	0.08	0.13	0.25	0.24	0.17	2.06	0.24	0.31
LOI	0.23	0.68	n.d.	n.d.	0.90	1.00	n.d.	n.d.
Total	99.89	99.69	99.49	98.96	99.87	100.36	100.0	100.0
Mg#	57.9	49.6	43.7	43.9	51.8	n.d.	56.0	43.9
Trace elements in ppm
Ba	60	53	372	581	266	724	n.d.	n.d.
Sr	810	180	457	569	208	n.d.	n.d.	n.d.
Zr	n.d.	34	293	187	n.d.	335	n.d.	n.d.
Y	n.d.	n.d.	67.6	31	n.d.	n.d.	n.d.	n.d.
Cr	40	n.d.	22	15	200	36.4	n.d.	n.d.
Ni	n.d.	n.d.	25	23	n.d.	n.d.	n.d.	n.d.
Sc	n.d.	n.d.	32	25.2	41	58.6	n.d.	n.d.

Whole-rock composition of gabbro (G) PP43, ST114 from [35] and TQ9a, TQ9b [this study]; gabbroic anorthosite (GA) P4 and oxide–apatite gabbro (OAG) P6 beneath anorthosite sheet from [23]; and amphibolite (AM) SF10, SF11 associated with anorthosite [this study]. NJH, New Jersey Highlands; NYHH, New York Hudson Highlands. Fe₂O₃^t, total Fe as Fe₂O₃; LOI, loss on ignition; n.d., not determined.

Megacrystic amphibolites in the New Jersey Highlands are also lithologically and geochemically similar to Mesoproterozoic mafic anorthosite (hornblende gabbro) dikes that are spatially associated with anorthosite in the Honey Brook Upland described by [22]. Based on the field relationships, the comparable geochemical compositions between the New Jersey and Adirondack amphibolites, the presence of unexolved feldspar megacrysts with low K contents, and the spatial association of both with anorthosite, we interpret megacrystic amphibolites in New Jersey to be coeval with AMCG magmatism in the Adirondack Highlands.

Megacrystic amphibolites in the New Jersey Highlands intrude ca. 1185 Ma granitoids and the available data strongly suggest they are geochemically equivalent to mafic intrusions elsewhere, as noted above, that are petrogenetically related to 1165–1140 Ma anorthosites. They may also be coeval with the alkaline to tholeiitic Abitibi dikes, Kingston dikes, and Robe Noire dikes in the Canadian Grenville Province that were emplaced at 1171–1141 Ma [16].

5.4. Petrogenesis and Tectonic Setting of Megacrystic Amphibolites

Megacrystic amphibolites have geochemical compositions that are characterized by Mg# of 40–51 and low abundances of MgO and transition elements (Cr and Ni), indicating they are relatively evolved compared to primary mantle melts. Their compositions are comparable to high-Al gabbro that is widely distributed in the New Jersey Highlands at the present erosion level (Table 4). Our results are thus consistent with the formation of the megacrystic amphibolites through fractionation of a source that was compositionally similar to high-Al-Fe gabbro. The available data also argue in favor of the petrogenetic relationship of these amphibolites to anorthosite bodies in the study area. On an AFM diagram (Figure 6), megacrystic amphibolite, gabbro, gabbroic anorthosite, and anorthosite from the study area fall along the same trend, strongly suggesting they may form a common liquid line of descent.

Mafic dikes that crosscut spatially associated anorthositic intrusions in the Laramie anorthosite complex in Wyoming [11], the Adirondack Highlands [12], and the Archean Bad Vermillion Lake anorthosite complex in Ontario, Canada [50], were emplaced late in the magmatic sequence. The dikes are interpreted as having formed as a differentiate of gabbro that was parental to anorthosite. The petrotectonic setting of these mafic dikes is consistent with the inferred field relationships of the megacrystic amphibolites in the New Jersey Highlands and their petrogenetic affinity to gabbro that suggest they were also emplaced late in the magmatic sequence.

Megacrystic amphibolites in the New Jersey Highlands clearly formed following the termination of subduction-related calc-alkaline magmatism at ca. 1.2 Ga and the emplacement of hastingsite and hedenbergite granitoids at ca. 1182–1188 Ma. The geochemical similarity of the megacrystic amphibolites in the study area suggests that they are comagmatic and were formed from a common enriched lithospheric mantle source. All of the samples reflect metasomatism of their source through the input of a subduction component that we propose is responsible for the slight enrichment in δ¹⁸O, and they plot above the mantle array on a diagram of Th/Yb vs. Ta/Yb (Figure 7a). Elements that preferentially partition into slab-derived fluid (Ba, Rb, La) versus subducted sediment, slab melt, or continental crust (Nb, Th, Ta) help identify subduction-related contributions to the mantle source [51]. On a diagram of Th/Nb vs. Ba/Th (Figure 7b), some samples of megacrystic amphibolite (TQ423-3, TQ423-4, EAS2) follow a trend of high Ba/Th, consistent with a contribution from slab-derived fluid. Samples BLR205, ST106, and DOV189 are displaced to higher values of Th (Table 3) and have low Nb/U (2.3–20) and Nb/La (0.07–0.70) and high Th/Ta (1.5–14) ratios, which suggests they assimilated continental crust during their emplacement. Trace element mobility during granulite facies metamorphism is a possibility, but we note that all samples show no evidence for fluid alteration or veining and, with the exception of Ba, interpretations derived from Figure 7 are based on REEs, HFSEs, and Th, which are generally considered robust during high-temperature metamorphism [52]. Overall, the major and trace element composition of the New Jersey Highlands megacrystic amphibolites is consistent with their derivation from a high-Fe-Al tholeiitic magma through partial melting of enriched, subduction-modified, lithospheric mantle in the spinel peridotite stability field followed by crystal fractionation of olivine and plagioclase.

Interpretations of the tectonic setting along the eastern Laurentian margin at ca. 1.2 Ga in the New Jersey Highlands [26,27] are in general agreement with those proposed for other eastern North American Grenvillian terranes and involve the termination of subduction-related calc-alkaline magmatism, convective thinning or delamination of overthickened crust, and replacement of delaminated lithospheric mantle by hot asthenosphere (e.g., [53,54]). As a result, mantle melts accumulated at the base of the crust where they underwent varying amounts of fractionation to produce intrusions of gabbroic to anorthositic composition (Figure 8). The absence of an OIB-like geochemical composition in the megacrystic amphibolites does not support their origin from a mantle plume. The megacrystic amphibolites are tentatively interpreted as having formed contemporaneously with anorthosite in the Highlands.

We present three possible tectonic scenarios for the formation of the megacrystic amphibolites. In the first scenario, the amount of advected heat following lithospheric delamination may have begun to diminish following final closure of the arc and back arc against the eastern Laurentian margin, resulting in only minor volumes of mafic magma underplating the lithospheric mantle and lower crust beneath the Highlands. The second scenario involves periodic tapping of this mafic magma by extensional, crustal-scale fractures that may have interrupted the extensive fractionation and plagioclase separation necessary to form more voluminous anorthosite intrusions. Instead, this process resulted in the emplacement of a meager volume of anorthosite as thin sills and amphibolites as thin dikes containing abundant megacrystic plagioclase. In the third scenario, ponding of mafic magmas would have differentiated to produce more voluminous anorthosite beneath the present-day Highlands. In this case, the sparse amounts of anorthosite emplaced as sills and megacrystic amphibolite emplaced as dikes exposed at the present erosion level would represent the higher structural levels of more voluminous intrusions preserved at depth. This is consistent with the interpretation that small anorthosite bodies in the study area were emplaced as thin sheets (sills) into the adjacent country rocks [23]. Scenario three is particularly attractive given the preservation of a wide variety of lithologies in the study area that are also found in the larger massif anorthosite complexes elsewhere. However, we acknowledge that all three scenarios are consistent with the available data and at this time we are unable to distinguish between them.

Figure 7. (a) Plot of megacrystic amphibolites (solid circles) on a diagram of Th/Yb vs. Ta/Yb modified from [51]. Arrow shows vector for subduction zone enrichment. Lower crust (LC), bulk crust (BC), and upper crust (UC) from [55]. GLOSS (global subducting sediment) from [56]. (b) Plot of Th/Nb vs. Ba/Th showing trends for the modification of a source from subduction-related slab fluid and slab melt/subducted sediment.

Figure 7. (a) Plot of megacrystic amphibolites (solid circles) on a diagram of Th/Yb vs. Ta/Yb modified from [51]. Arrow shows vector for subduction zone enrichment. Lower crust (LC), bulk crust (BC), and upper crust (UC) from [55]. GLOSS (global subducting sediment) from [56]. (b) Plot of Th/Nb vs. Ba/Th showing trends for the modification of a source from subduction-related slab fluid and slab melt/subducted sediment.

The concordance of NE-striking metamorphic foliation in the megacrystic amphibolites with that in bounding country rocks contrasts with the east–west trend of the amphibolites. The concordance of this high-strain metamorphic fabric indicates that the megacrystic amphibolites and country rocks were metamorphosed together during Ottawan orogenesis at ca. 1045 Ma. The east–west trend is therefore consistent with the structural control of amphibolite emplacement in an overall NE-striking crustal-scale extensional fault system that developed a subordinate component of transtensional, right-lateral strike slip following the termination of subduction.

Figure 8. Geodynamic model (modified from [57]) showing late- and post-Elzevirian events in the New Jersey Highlands between ca. 1200 and 1130 Ma. Upper panel shows the termination of westward subduction and collision of the Losee arc at ca. 1200 Ma. Lower panel shows minor extension and lithospheric delamination at ca. 1160–1130 Ma causing asthenospheric upwelling that provides the heat source for partial melting of subduction-modified lithospheric mantle. High–Al–Fe magmas are produced and then pond in the lower crust, crystallize, and are emplaced as megacrystic mafic dikes. The dikes are then deformed and metamorphosed during the Ottawan Orogeny at ca. 1045 Ma.

Figure 8. Geodynamic model (modified from [57]) showing late- and post-Elzevirian events in the New Jersey Highlands between ca. 1200 and 1130 Ma. Upper panel shows the termination of westward subduction and collision of the Losee arc at ca. 1200 Ma. Lower panel shows minor extension and lithospheric delamination at ca. 1160–1130 Ma causing asthenospheric upwelling that provides the heat source for partial melting of subduction-modified lithospheric mantle. High–Al–Fe magmas are produced and then pond in the lower crust, crystallize, and are emplaced as megacrystic mafic dikes. The dikes are then deformed and metamorphosed during the Ottawan Orogeny at ca. 1045 Ma.

6. Conclusions

Penetratively deformed megacrystic amphibolite bodies have been identified from several locations in the New Jersey Highlands. The megacrystic amphibolites form conformable layers 0.5 to 2 m thick, have sharp, conformable contacts against Mesoproterozoic country rocks, and are best interpreted as metamorphosed dikes that contained large plagioclase megacrysts up to 13 cm long. The megacrystic amphibolites analyzed in this study have whole-rock major and trace element geochemical compositions characterized by high Fe and Al and enrichment of LREEs (La/Yb_N = 1.73–10.69) and high LILE/HFSE ratios (e.g., Ba/Nb = 24–211) that are similar to those of gabbroic and anorthositic rocks from the Adirondack Highlands and the Canadian Grenville Province. The geochemical composition of megacrystic amphibolites is consistent with crystal fractionation from a high-Al-Fe tholeiitic primary magma that formed by partial melting of an enriched, subduction-modified, lithospheric mantle source in the spinel peridotite stability field. The geochemical data presented here support a petrogenetic model that involves production of primary, mantle-derived mafic magmas produced by post-orogenic thinning of the continental lithosphere and subsequent ponding and crystal fractionation in crustal magma chambers to produce the high-Al-Fe magma with plagioclase megacrysts. Periodic tapping of these magma chambers by extensional, crustal-scale fractures and/or small amounts of magma production during lithospheric thinning could explain the lack of abundant anorthosite in the New Jersey Highlands. Timing of emplacement of the megacrystic amphibolite protolith in the New Jersey Highlands is broadly constrained to be between 1160 and 1130 Ma, which provides support for a coeval relationship with other pre-Ottawan AMCG complexes elsewhere in the Grenville Province.